Recently I did an interesting discovery: the work that I did during my PhD (1996 – 2000) might be at the basis of a radical new technology to produce anodes of Li-ion batteries, anodes based on silicon nanowires instead of graphite. The Dutch branch of the company Meyer-Burger (Eindhoven) has sold a large tool to the silicon valley-based company Amprius. This tool appears to use the so-called expanding thermal plasma (ETP) method that I worked on during my PhD. For my dissertation I studied the high-rate deposition of amorphous silicon films for the application in silicon thin solar cells. However this method never obtained sufficiently interest from industry and in the meantime silicon thin films are not considered a viable alternative anymore for crystalline silicon solar cells. However, now my work seems to be of relevance for a new type of Li-ion batteries.
Let me first give some more background information: as said, during my PhD I studied the deposition of amorphous silicon thin films by the ETP method. This is one particular configuration of plasma-enhanced chemical vapor deposition (PECVD). The unique selling point of the ETP method at that time was the high-deposition rate that could be achieved: silicon films could be deposited at a rate of ~10 nm/s which was a factor of 10-100 faster than with conventional PECVD methods at that time. For solar cells, this was a strong selling point as “costs” is everything in the field of photovoltaics. However, as mentioned, the method never made it for the deposition of silicon film solar cells and a couple of years ago the silicon thin film solar cell technology itself went down the drain. But there was also a side project during my PhD. I also did some work on high rate deposition of silicon nitride thin films. Although this was only a small effort during my PhD, this work became much more important and relevant in the end. During my postdoc and the early years as an assistant professor, the ETP method for silicon nitride deposition was scaled up and commercialized for high volume manufacturing of crystalline silicon solar cells. The silicon nitride was applied as an antireflection coating and passivation layer on such cells. The Dutch company OTB Engineering – later OTB Solar – developed a high throughput system – the DEPx – based on the ETP method. Over the years, this became their main product of the company and they sold 100+ systems.
Well, I want to come back on this silicon nitride system in another blog at some point but let’s now return to the preparation of silicon films by the ETP method. It is important to realize that the company OTB Solar was acquired by the German company Roth & Rau, which itself was acquired by the Swiss company Meyer-Burger soon afterwards. So it is basically the same company that now started using the ETP for the fast deposition of silicon thin films, this time for battery applications.
What is exactly known about this? There has been a press release by Amprius stating that “Amprius, a leading manufacturer of high-energy lithium-ion batteries, has demonstrated a revolutionary new tool for high-volume manufacturing. The new tool, a first-of-its-kind system for inline, continuous, and roll-to-roll production of three-dimensional silicon nanowire anodes, will enable Amprius to scale manufacturing and deliver lightweight and long-lasting batteries for unmanned vehicles, wearable technologies, and electric vehicles.” More specifically about the tool it mentions “The new manufacturing tool, which Amprius developed in partnership with the Dutch firm Meyer Burger, uses a multi-step Chemical Vapor Deposition (CVD) process to produce Amprius’s silicon nanowire anodes.” According the press release Amprius believes that with silicon anodes significantly higher energies per unit volume and per unit weight can be achieved than by the carbon-anode designs commercially available today.
Also the Charged Electric Vehicles Magazine covered this topic and this magazine even provided a photograph of the tool. To me, as an insider, this photograph made it very clear that the ETP method is used in the tool. On the right side of the picture, one can see three power supplies that are used in welding. They are located on top of the vacuum system, right before chamber where the foil is unwinded (or winded up). It even seems that there are four power supplies, with one being shielded by blue beam. These welding power supplies are being used as power sources for the ETP sources used. We can therefore conclude that the system contains 3 or 4 ETP sources. The sources are most probably located in the middle part of the system. This part of the vacuum chamber is slight higher and this is necessary to have the plasma beams expanding from the sources. The length of this part is relatively short as the ETP typically acts as a relatively short line source with the substrate or foil moving underneath it. The deposition rate is very high such that not many sources are needed. The large parts of the vacuum chamber before and after the part with the ETP sources can be substrate heating and cooling zones as well as pre- and post-treatment chambers. Note that the Amprius press release talks about a multi-step CVD tool.
On the internet I also found some interesting Department of Energy (DoE) reports when searching for Amprius and its founder Yi Cui, a professor at Stanford University. One report (Advanced High-Performance Batteries for Electric Vehicle (EV) Applications; June 2016) shows a photograph of a web which is transported through a tool. The associated text mentions that the nanowire anodes are produced at both sides of the web so perhaps this means that the deposition of the thin silicon films takes place from the top and from the bottom of the tool.
Finally, one also can find some interesting patent applications from Amprius. For example, the US patent application US 2015/0325852 A1 with the title “STRUCTURALLY CONTROLLED DEPOSITION OF SILICON ONTO NANOWIRES”. The patent application yields some more details about the silicon nanowires, e.g., the figure shown below. This figure shows a non-conformal silicon layer deposited by PECVD over a nanowire template. On top of this layer a second layer is deposited by thermal CVD. Claim 17 is very interesting: “The method of claim 16, wherein the PECVD method is an expanding thermal plasma (ETP) method”!
I guess it is quite clear why I’m so excited about this “discovery”! Perhaps my PhD work will be applied in industrial manufacturing in the end, not in manufacturing of solar cells but of batteries. Battery technology is a field that needs even more advancement than solar cells! And this might just be the beginning. The aforementioned patent also reads: “The conformal, dense silicon layers may be deposited by a method as atomic layer deposition (ALD) instead of or in addition to thermal CVD”.
Some slides with more details on the silicon nanowire anodes can be downloaded here.